1. Forum
  2. >
  3. Topic: Duolingo
  4. >
  5. Are people translating enough?


Are people translating enough?

I've been on Duolingo for nearly 4 months now and I've noticed a tendency in myself and others to avoid translating a lot of sentences and instead opting to test out of skills. Duolingo provides a ton of motivation to rack up points, skills, and levels, but there isn't as much motivation to do translations. They are difficult, slow, and messy, and they don't give you as many points for the effort expended as lessons and tests do.

Perhaps translating is not so important in the lower levels of Duolingo, but even people at higher levels are not translating much. At level 12 I rarely find a sentence that is translated by anyone other than the duobot. I wonder what happens after you get through all the levels? Where will the motivation come from to give back to Duolingo and translate more sentences when there aren't any more points and skills and levels to achieve? I think the translation side of Duolingo could do with more gamification. Perhaps translations should be worth more points, and perhaps the tests to master skills should be more difficult. Or perhaps there should be a minimum number of translations required to master a skill. Getting people in the habit of translating for every skill might be a good thing. What does everyone else think?

May 7, 2012



Totally agreed. We're working on improving this as we speak!

[deactivated user]

    1) Overall I agree, it seems people are not doing a lot of translations.

    2) I think the suggestion to require translating between 3 - 10 sentences per level (depending on the level) before completing the level is a good one.

    3) As you mentioned translation is not easy because most of the texts use vocabulary that learners are not familiar with so it feels painfully slow and awkward. Lessons are comparatively easier and provide an immediate reward. One possibility is to include translation DIRECTLY into the lessons! Say you have to get above 50% on 3 translations to complete a lesson.

    4) Another issue is that many of the articles for translation are about topics like fashion and celebrities that many people find dull. I actually didn't mind this at first but it gets pretty boring after the 20th time. On the other hand, just yesterday I translated part of an article about Karl Marx and I enjoyed it quite a lot even though it was full of long sentences, simply because the content was very interesting to me. Perhaps Duo should allow customization of the topics you want to translate (mine would be politics and history because I usually learn interesting stuff while translating them).

    But in general they are still tweaking the system, I am sure they will figure something out.


    Thanks, you all bring up an excellent point about some of the boring translation articles. I also vote for no more fashion articles! They would probably be safer with travel articles, since the people here are obviously going to be interested in other cultures. I wonder how hard it would be for them to match people's interests with web articles? At least they could offer more than two options


    I think it would help to show people how good their translations are, which would give them motivation to translate better and more. It's already doing this to some extent, in that it shows you on the "Translations" page how many upvotes you've received. However, perhaps using upvotes/downvotes/number of translations/number of skill points etc. you could generate an overall translator's score, and a person would want to maximize that score.

    Another idea is to make a game out of it by adding accomplishments/badges. For example, you get a special badge for doing 1000 translations, receiving 1000 upvotes, or for translating 100 sentences in a single day. If you really want to motivate people, perhaps they could unlock things with their completed translations? I don't know how much partnering you can do, but what if a certain amount of translation points could be redeemed for a tangible prize; something small, like a free iTunes song download etc. Obviously that's a pretty involved idea, but it's the kind of thing you have to do if you want people to keep using the program long term.


    @amichail That's an interesting point. Maybe presenting translation articles that are inherently interesting to the user is the way to go. I read a book by Kato Lomb, a famous Hungarian polyglot, and her method for learning languages was to read books on topics she was vitally interested in. Wanting to know what they contained gave her the motivation to keep slogging through. (The book is a great read. You can download it for free from here: http://tesl-ej.org/ej45/fr1abs.html )


    It's seems the book is no longer available. Do you have any other source? Thanks.


    Thanks a lot! And sorry for needlessly asking this. It's not a habit of mine to do so, but it seemed a rather obscure subject, and subconsciously I didn't think it was worth the search, I guess.


    I think you've made some valid points. I find I waffle between doing just lessons and tests, then doing just translations. Sometimes the translations offered are outside of my interests or skills set so I will take a stab at them but then lose interest and go back to the lessons. If I'm offered translations that are on topics of interest then that's all will I do. I'm getting close to completion on the levels but I don't see myself losing interest in Duolingo any time soon - it's too much fun (almost addictive) - and with the strong sense of community that is developing I will choose this over any "social media" platform any day of the week!


    My motivation with Duolingo is to open more skills. While learning Spanish, I first opened everything (by reaching the middle of the skill) and then started completing all the skills to the "master" degree one-by-one. When I was done with Spanish I started French. I had some motivation to do translations, but soon enough I realized it gets me nowhere. Most of the time I had to "translate" a football player's name and birth date, or extremely complicated sentences without any added value to the learning curve. So after mastering all Spanish I just started learning the French. I know that this is kind of unfair to Duolingo's goals, but I'd love to do translations if they get me somewhere in terms of learning the language or at least having a sense of achievement. How about a chart of the translation-contributors-of-the-week/month/year? How about not opening the next skill before having so-many translations? How about giving shorter sentences, while showing the surrounding context? This part surely need to improve in the motivation and added value parts.


    I doubt that making translations a requirement will motivate people to do them. There are different reasons for not doing them and some of them have been expressed here in this thread. Let us take the point of not making a translation because the subjects are not interesting to you and you get bored. What would happen if you are forced to do the translations now? You not only would still be bored but probably also get frustrated or angry because you now have to do those boring translations if you want to advance. That will surely not help in motivating you to do more translations. If I would be forced to do some things I don't want to or don't like, I probably would think about leaving. If you force someone to do something you don't like, there always needs to be enough incentive to "endure" those tasks.

    And now think of a different solution. Removing as much of the reasons for not making translations and/or giving the users as much incentive as possible to do translations. In my eyes that would be a better solution then the first one. There might be still some users that will not do translations but who says that everyone needs to do them.


    To a great extent I agree with you rweba that the articles can be boring and a little frivolous; however I find that translating helps teach how the language is actually used. I took Spanish for a few years during school, but until I started translating it here, I had no idea how discursive the language is. I might not give a whit about what Jessica Alba wore to some premiere, but learning the syntax is actually very helpful. That said, there are plenty of interesting articles out there. Just tonight, I translated the review of a re-issued book authored by a late Franco-Polish artist and Roman Polanski collaborator. The book itself isn't available in English, so I would not have otherwise known of its existence.


    If they are available, reading good translations of the sentences is also very helpful.


    I do find that as my Spanish is not very good that some of the articles are just too difficult for the lesson level. I think if I was a complete beginner I would find translating a bit intimidating. I agree that the subject range is a bit limited. The French site seems to get more junior Wicki which I think are good for beginners. Also with some lessons I want to do the practice exercises rather than the translations as I want to improve my grammar. A couple of times I have got the grammar points for translating and lost the exercises without being able to see what they covered. I think it would be good to be able to have a space to be able to discuss the articles and any translation problems with them.


    For me, too, the biggest problem with the translations are the topics. I don't mind spending half an hour to understand a complex construction of nested sentences if I'm actually interested in the topic. But if I have to spend the same amount of time and effort on an article about fashion, interior design or just gossip, my motivation drops down dramatically...


    I learn from two sources: duolingo & spanishdict.com on itunes to learn spanish, its also free. I also find that the translation choices have nothing to do with what level of learning we are in. perhaps this could be a reason not to many folks try to translate.


    One would expect higher quality translations if people are not forced (or too highly encouraged) to do them.


    All excellent points, however, I would just like to say that some of us are not that smart and enjoy the "mindless" topics such as fashion and interior design!


    @ClassicBookworm: Thanks for the suggestion. Another great resource - despite its age - for aspiring polyglots is Bodmers 'The loom of language'. However, I don't know whether it is available for free somewhere...


    I can understand some of the reasons mentioned here for not doing translations. Doing translations only for topics that are not of interest to someone can be boring. But with the change of the layout lately, I seldom find translations that I find really boring, as there are not only the translations in the lessons but also the ones under the heading "Translations". And having a shot at a translation of a topic outside ones range of interests can be interesting and helpful sometimes. What probably would be interesting, although sometimes a bit harder, would be lyrics.

    Also, the thought of translating something as a "complete" newbie can be really daunting. But I think that one should have a try on them nevertheless. Most translations have at least some sentences with lower difficulty levels, that can be translated as a beginner. Even the more complicated sentences can be fun, even if I don't get the meaning completely or at all. I just try to see them as puzzles or riddles, trying to solve them or in this case finding an English sentence that seems to make sense. The only thing I really miss here, is the possibility to review how others have rated my translation. Not necessarily who rated me bad or good but how my translation is seen by others. Sadly I can't see that at the moment.

    I also wouldn't make translating a requirement. Some people might not like that and I couldn't blame them. Learning languages (and learning in general) should be fun. I think that forcing users to do something, is not the way to do that. I agree with amichail there. The buzzword "Gamification", like ClassicBookworm said, could probably give us some pointers. The goal would be to make other users interested in the translations through showing them (or adding) benefits and removing resistances (the reasons why they shy from them).


    What's most discouraging to me about doing translations is when I'm the first to translate an article and give a great and natural-sounding translation which is then denied because it doesn't match DuoBot's very bad translation. What we really need is a way to feel like a team in translating each article. What if each article was offered to at least 10 people and those people had to contribute to that one before moving on to the next? Then they could work together and vote on the best translation, and discuss it, then get points as a group, with whoever offered the best translation getting more points? Some way to get a team effort like this would help build the community and also result in much better translations. I know this is already partly implemented with being able to rate other people's sentences, but no points are offered yet once you have given your own translation.


    If I am having difficulty with a skill, say Infinitives, then I have found that doing as many translations as I can really helps me learn and it has the added bonus of directly adding points to the level I am having trouble with.

    Apart from points and levels though, the intention of DuoLingo is to translate "the web". So, I do translate, regardless of how personally interested I am in a subject. It is my pleasure to find I am enjoying some articles I would not usually choose. If sometimes the articles are boring, then that is a small price to pay for such a great program.

    However, since people seem to be motivated by points, perhaps there is another way to engage them such as ranking them overall in the number of translations done (not just those they follow) or having a separate translation points system?

    Or simply make it mandatory to translate a certain number of sentences before new levels are made available.


    Interesting, that so many people seem to not understand the percentage we receive after submitting a translation. I know it can be disappointing, if the percentage is low, especially for newbies in that language. But even if it is zero (which I had a few times), that only means that none of the previous translations are like yours. Yes, that can mean that your translation is completely wrong but also that the others are completely wrong (which I had as well). Maybe there should be a better way to display such a value, as I personally use that value as a sign, if I need to check the other translations or can skip that part (100% agree with my answer).

    Teaming users together? OK, I can see the effect of the feedback in a team but with too many problems. Giving feedback for a translation is very helpful and we already have the possibility to suggest edits for the translation of others where we also can explain why we think our solution is better. And we even can get or give feedback to those edits, which sometimes leads to a shorter or longer discussion. Joining users together to teams would be problematic in this case as we would need a way to choose the users which would build a team, there would also be the problem of coordinating those users so that the progress of one user is not hindered by the team. I think, a better way would be improving the possibilities for feedback, e.g. making the "suggest edit" link not only visible when the mouse is hoovering over the sentence.

    @Shumulik - Have you seen the top lists under the heading "Translations"? There you'll find the top translators for the last 24 hours and the past week. When was the last time you tried to translate? Duolingo lately changed some parts of the website and you even find complete wikipedia articles awaiting translation. That should surely be something that would add value to your learning curve.


    My proposition for this problem is the following: There are points that you can earn form tests/reviews/testing out skills, and points from translations. Right now, to pass a skill, you only need points that are equivalent of doing all lessons. Maybe add another bar, with points that can be only filled by tanslations? This way, for example, to pass "Verbs 1" you would need 120 points from lessons, and 40 points from translations. I think it would motivate people to translate on a regular basis, and could even be an improvement on learning process.


    Translating a random wikipedia page is indeed tedious.

    It would make more sense to allow users to translate the text from arbitrary web pages in the language they're learning. I can do this already by correcting the Google webpage translations, but I don't expect my machine-translated results to improve soon or ever, so it feels like wasted effort. They don't even apply my corrections elsewhere on the same page.


    I find that the translations tend to require a skill level greater than would be expected from a beginner to an unfamiliar language. To translate from German to English can be difficult because the context of an English phrase can change depending on how the sentence is arranged. That being said, I choose to translate based on context rather than sentence structure because the two languages, German and English, do not necessarily translate cleanly. I will take the lower score for a more accurate translation. What's the point in meeting a "bot's" expectation if the end translation is not correct?


    I've primarily been doing the lessons (if I feel I need them), and then taking the "opt out" test, and THEN doing translations. Call me silly, but I like having the nice, uniform golden color.


    @angiedaytripper That's a great suggestion about having a place to discuss the articles and their translation problems. I hope Duolingo is reading this!


    I translated an article on Formula 1 racing today, at one stage every second task was a number, like 2. or 4. and I got six points for "translating" each number. Needless to say I soon got bored. Generally the phrases have been okay though. I wonder if it would be worth giving bonus points if you do the exercise without hovering over words for translation? Also I would like some guidance as to whether we should be doing a literal translation or one that uses more colloquial expressions (I prefer the latter).


    I actually find that you receive to many points for a translation as I don't wish to progress through the levels before I feel comfortable with how I'm understanding them.

    Having to translate a certain amount of sentences prior to proceeding to the next level is a good idea. With that said, I think once one gets to a certain level (as an incentive) they should be able to choose the subjects they would like to translate. Different subjects could then be selected as a person moves up the learning ladder.


    The translation are the whole point of the learning for me. It gives me purpose, and it is in some way what it must be like learning the language through immersion in the language context. Th translations are fun and there are plenty of choices, for me at least. If I don't like them, they change after a fairly short time, so I generally find something to engage me.

    I also like the fact that the translations offer me sites that I would not stumble across myself. Given the unfamiliarity of some of the material, a way of discussing the specifics of translating a particular page at the point of translation would be very helpful.

    The choice between wholly a literal translation and one more in keeping with the language being translated into is fraught. Literal translations still sound 'foreign' in many cases, but a more fluent one receives a much lower score, because of comparison with duobot.


    In my opinion forcing beginners (or people who just learned a particular subject) to translate is not a good idea. You will end up with tons of bad quality translations or even worst ... with google translate answers. Why would anybody do that? Because he doesn't like the subject, he feels confident, etc... and he wishes to go on. I think that a good place for "forced" translations are the ending tests and "magic shortcuts". If someone is good enough to master the chapter, he may do 5 translations. Also, as stated before, some translations should came from lower levels, so an experienced user gives good translations to beginners to practice their skills. I would like, also the possibility to mark(highlight) the words I didn't like when giving a bad rate to someones translation. I wish I could see the parts of my translations others didn't like and their (better or not) translations. Leaving a place for discussion ("My word is better because...") would end up in unneeded arguing, but word marking gives the needed information to the author. By the way thank You for this wonderful tool:)


    My motivation with Duolingo is to open more skills. While learning Spanish, I first opened everything (by reaching the middle of the skill) and then started completing all the skills to the "master" degree one-by-one. When I was done with Spanish I started French. I had some motivation to do translations, but soon enough I realized it gets me nowhere. Most of the time I had to "translate" a football player's name and birth date, or extremely complicated sentences without any added value to the learning curve. So after mastering all Spanish I just started learning the French. I know that this is kind of unfair to Duolingo's goals, but I'd love to do translations if they get me somewhere in terms of learning the language or at least having a sense of achievement. How about a chart of the translation-contributors-of-the-week/month/year? How about not opening the next skill before having so-many translations? How about giving shorter sentences, while showing the surrounding context? This part surely need to improve in the motivation and added value parts.


    Would people really object so much if they had to do translations to master a skill? After all it's the basis of the Duolingo idea--learn a language by translating the web. Those who only want to learn without giving anything back are not the sort of people Duolingo wants on the system anyway, so there is no harm if they leave. If translations were made mandatory and that was the only system future users knew, I think they would just accept it as part of the deal.


    Personally, as a beginner I've found that translating by itself doesn't really help my language skills much. I have learned a few things, and it is becoming more meaningful as I've gotten further along, but it doesn't help nearly as much as the drills. I hate to admit it, but I find that the point-based system is quite motivating (as many points translating one sentence as by answering lots of questions). The new goal of 100 points a day helps with that.

    The initial poster seems to be referencing the end-of-unit translations. For me, I find them rarely interesting, so I've personally mostly abandoned them. To find topics that interest me, I depend on the "Translations" tab at the top of the screen and on my "Stream" to see what the people I follow found interesting enough to translate.


    @CB Just reviewing this thread and want to thank you for the link to Kato Lomb's book. I'm half way through it already. Wonder what she would have made of YouTube? @LeoSynapse, I agree with you, I want my translations to carry the meaning of the original and I will happily change the syntax and use colloquial expressions to get the meaning across, even if it means using different verbs, nouns etc. I think the key to good translation is to ask yourself "what does the writer or speaker mean?" and then explain it to your audience using every linguistic resource at your disposition.


    @wataya Thanks for the tip on The Loom of Language. My library has it.


    @Bill I'm sure Kato Lomb would have been absolutely thrilled by the web and learned twice as many languages in half the time! For any enterprising person there is really no barrier (other than internet access) to learning any language now.

    Learn a language in just 5 minutes a day. For free.